NexPCB Blog

Wearable device manufacturing: what a working prototype still has to prove

Written by NexPCB United | Nov 16, 2026, 3:21:29 AM

A working wearable sample proves the design can run once. It does not prove the build can be repeated. Between the two, three things have to be settled: which conditions change the build, how the work moves from consulting to volume, and what a pilot has to show. At NexPCB we have worked on hardware since 2005 and have completed more than 600 products (About page). Our wearable solutions cover the wider service.

Constraints that change the build

Three conditions change how a wearable is built.

Skin contact and wear time put the first risk on the materials and the seal. Artificial sweat, salt spray, waterproofing, ultraviolet exposure, and high and low temperature are the tests that show whether they hold.

Size forces the electronics and the enclosure to be designed together. A board that fits the sample can still miss the opening, the seal, or the connector once the housing is closed, so we work the two in parallel.

Heat and the cell raise three separate questions, and each needs its own answer. In use, heat and the cell set the power path: how heat leaves the device and how the cell is supplied. On the finished device, temperature cycling and drop show whether the build survives handling. Shipping the cell is a third question. UN 38.3 is the lithium-cell transport test used for that. A pass there says nothing about how the finished wearable behaves hot or dropped.

How wearable device manufacturing proceeds

We run the work in four stages: consulting, EVT/DVT, pilot production (NPI), and volume. Each stage ends when something concrete can move forward. The files and tests behind that step are set per project. A technical proof of concept comes in when the project needs one.

StageWhat this stage settlesOur service line
ConsultingStart from the product definition, the intended use, the current concept, and the business goal. We weigh the architecture and the development and supply-chain risks, and leave a budget and milestone plan with a verification target and a clear route into the next stage.Innovation Consulting
EVT/DVTThe sample, design files, bill of materials, test requirements, and quality target go into a DFX review and the engineering checks. What leaves this stage is a manufacturing package, a result against the agreed tests, and the revision the pilot will use. The control book firms up with that package.Engineering Acceleration (NTI)
Pilot (NPI)With the design and material revision locked for the pilot, plus the test and quality requirements and the batch target, we move tooling, first-article inspection, assembly, and test setup into the agreed build. What has to leave with the batch is that record, and a written disposition for anything still open.Production Delivery (NPI), after NTI
VolumeOnce the approved package and sample are in, with the order, the demand outlook, and the delivery plan, we run the batch under quality control and keep supply and repeat delivery moving. Later material, supplier, or process changes stay in view, so acceptance includes batch consistency and a trace of what changed.Production Delivery (NPI) and Product Lifecycle Management

The control book is what ties the early stages to the pilot. In engineering acceleration we take a design from prototype iteration into a verified pilot product and prepare that book. In production delivery we start from it before tooling, and first-article inspection comes before a full batch. Later supply, inventory, and change control sit in product lifecycle management.

What a wearable pilot has to prove

A pilot is where we find out whether the manufacturing path repeats. Another round of samples cannot answer that.

Our optical-window case shows why. The first pilot batch that shipped showed weaker sensor signals than the prototype batch. We held the remaining units and opened an investigation.

We collected device-level data and compared two optical-window batches on the same device. A coating-process change at the window supplier had changed infrared transmittance in the working wavelength range. The full investigation and the change control around it are in How an Optical Health Wearable Moved From 20 Prototypes to 250 Pilot Units.

On the next program of this kind, we will fix the window’s working wavelength, the transmittance requirement, and the batch test report before the pilot, and put that material check on the pilot list. The production checks that follow a pilot are in Eight checks that decide if your wearable is ready for mass production.

Frequently asked questions

After a prototype works, what still has to be proved?

That the build repeats. The sample runs, but the program may still be in consulting or EVT/DVT. What remains is the pilot record, then the batch record. A sample that works on the bench closes neither.

What does each manufacturing stage have to prove?

Consulting closes when the verification target, the main risks, and the route into the next stage are clear. EVT/DVT closes on a result against the agreed tests and the revision the pilot will use. The pilot closes on a first-article and test record, with a disposition for anything still open. Volume closes on batch acceptance and a trace of later changes.

What to bring to the first conversation

Tell us the current stage, the intended use, and the one engineering or manufacturing issue you need to resolve. If a sample, a design file, or a test note already exists, say what it is. We then use the stage table to decide what still has to be demonstrated before a pilot build: design verification, a critical material check, the manufacturing package, or the test method.

Share detailed design files once a non-disclosure agreement is in place. Start at our project inquiry form.